gecko/layout/reftests/canvas/reftest.list

== default-size.html default-size-ref.html
== size-1.html size-1-ref.html

== empty-transaction-1.html empty-transaction-1-ref.html

== image-rendering-test.html image-rendering-ref.html
== image-shadow.html image-shadow-ref.html

asserts-if(cocoaWidget,0-2) == size-change-1.html size-change-1-ref.html

!= text-ltr-left.html text-blank.html
!= text-ltr-right.html text-blank.html
!= text-rtl-left.html text-blank.html
!= text-rtl-right.html text-blank.html

== text-ltr-start.html text-ltr-left.html
== text-ltr-end.html text-ltr-right.html
!= text-ltr-left.html text-ltr-right.html
== text-rtl-start.html text-rtl-right.html
== text-rtl-end.html text-rtl-left.html
!= text-rtl-left.html text-rtl-right.html

== text-ltr-left.html text-rtl-left.html

== text-ltr-alignment-test.html text-ltr-alignment-ref.html
== text-rtl-alignment-test.html text-rtl-alignment-ref.html

== text-horzline-with-bottom.html text-horzline.html
== text-horzline-with-top.html text-horzline.html

!= text-big-stroke.html text-blank.html
!= text-big-stroke.html text-big-fill.html

== text-context-state-test.html text-context-state-ref.html
== text-font-inherit.html text-big-fill.html
== text-space-replace-test.html text-space-replace-ref.html

== text-no-frame-test.html text-no-frame-ref.html
== text-no-frame-2-test.html text-not-in-doc-ref.html
== text-not-in-doc-test.html text-not-in-doc-ref.html

== text-bidi-ltr-test.html text-bidi-ltr-ref.html
!= text-bidi-ltr-test.html text-bidi-ltr-notref.html # for bug 698185
== text-bidi-rtl-test.html text-bidi-rtl-ref.html

fails-if(Android) != text-font-lang.html text-font-lang-notref.html

== text-measure.html text-measure-ref.html

== strokeText-path.html strokeText-path-ref.html

# check that emoji character renders as something non-blank (for Apple Color Emoji font, bug 715798)
random-if(!cocoaWidget) fails-if(OSX==10.6) random-if(OSX==10.7) != text-emoji.html text-emoji-notref.html

# azure quartz uses CGDrawLinearGradient instead of DrawShading
# so we have less control over degenerate behaviour as tested by this
# test
fails-if(azureQuartz) == linear-gradient-1a.html linear-gradient-1-ref.html

# this passes with cairo on 10.7 and 10.8 but not with azure for reasons unknown
fails-if(OSX==10.6||(azureQuartz&&(OSX==10.7||OSX==10.8))) == linear-gradient-1b.html linear-gradient-1-ref.html

== zero-dimensions.html zero-dimensions-ref.html

== evenodd-fill-sanity.html data:text/html,<body>Pass
!= evenodd-fill-1.html nonzero-fill-1.html
== evenodd-fill-1.html evenodd-fill-ref.html
== evenodd-fill-2.html evenodd-fill-ref.html
== evenodd-fill-3.html nonzero-fill-2.html

== dash-sanity.html data:text/html,<body>Pass
fuzzy-if(azureSkia,9,470) random-if(Android) == dash-1.html dash-1-ref.svg  # Bug 668412 (really is android-specific, not IPC-specific)

== ctm-sanity.html data:text/html,<body>Pass
fails == ctm-singular-sanity.html data:text/html,<body>Pass  # Bug 612033
== ctm-1.html ctm-1-ref.html

fails-if(OSX==10.6) == 672646-alpha-radial-gradient.html 672646-alpha-radial-gradient-ref.html # Bug 673333
== 674003-alpha-radial-gradient-superlum.html 674003-alpha-radial-gradient-superlum-ref.html

!= 693610-1.html 693610-1-notref.html # bug 693610: multiple glyph runs should not be overprinted

== 726951-shadow-clips.html 726951-shadow-clips-ref.html

== transformed-clip.html transformed-clip-ref.html
== transformed-gradient.html transformed-gradient-ref.html
== transformed-path.html transformed-path.html

== 749467-1.html 749467-1-ref.html

# You get a little bit of rounding fuzz on OSX from transforming the paths between user space and device space
fuzzy-if(azureQuartz,2,113) fuzzy-if(d2d,12,17) == 784573-1.html 784573-1-ref.html

== 802658-1.html 802658-1-ref.html
Add reftest for default width and height of <canvas> (bug 291285, bug 293225) 2009-08-12 20:07:02 -07:00			`== default-size.html default-size-ref.html`
Bug 691354 - Enable test fix by changes. r=jmuizelaar 2011-10-17 13:09:03 -07:00			`== size-1.html size-1-ref.html`
Add reftest for default width and height of <canvas> (bug 291285, bug 293225) 2009-08-12 20:07:02 -07:00
Bug 787623. Call MarkContextClean immediately if we're not going to receive a DidTransactionCallback. r=joedrew --HG-- extra : rebase_source : 459c517fba4b6c2e6e6d95200aebec9369c0edb6 2012-09-05 03:28:23 -07:00			`== empty-transaction-1.html empty-transaction-1-ref.html`

Bug 486936 - image-rendering style should apply to canvas element. r=vladimir 2009-06-17 13:31:18 -07:00			`== image-rendering-test.html image-rendering-ref.html`
Bug 662898. Test 2011-06-12 18:42:48 -07:00			`== image-shadow.html image-shadow-ref.html`
Bug 486936 - image-rendering style should apply to canvas element. r=vladimir 2009-06-17 13:31:18 -07:00
Bug 581592 - allow for a couple of asserts in size-change-1.html on Mac, r=dbaron, a=orange --HG-- extra : rebase_source : 6d7ab1eeb272786e75fa57521d3a5c0f698fb88e 2010-08-29 21:51:56 -07:00			`asserts-if(cocoaWidget,0-2) == size-change-1.html size-change-1-ref.html`
Bug 564991. Part 2: Create retained layer API: methods on FrameLayerBuilder that frames can use to access, reuse and invalidate retained layers. r=mats,sr=vlad 2010-07-15 14:07:46 -07:00
Canvas text spec implementation (bug 436904) 2008-06-10 16:16:59 -07:00			`!= text-ltr-left.html text-blank.html`
			`!= text-ltr-right.html text-blank.html`
			`!= text-rtl-left.html text-blank.html`
			`!= text-rtl-right.html text-blank.html`

			`== text-ltr-start.html text-ltr-left.html`
			`== text-ltr-end.html text-ltr-right.html`
			`!= text-ltr-left.html text-ltr-right.html`
			`== text-rtl-start.html text-rtl-right.html`
			`== text-rtl-end.html text-rtl-left.html`
			`!= text-rtl-left.html text-rtl-right.html`

Canvas routines draw right-to-left text backwards - bug 402276 r=smontagu sr=roc 2008-07-18 11:29:06 -07:00			`== text-ltr-left.html text-rtl-left.html`
Canvas text spec implementation (bug 436904) 2008-06-10 16:16:59 -07:00
			`== text-ltr-alignment-test.html text-ltr-alignment-ref.html`
			`== text-rtl-alignment-test.html text-rtl-alignment-ref.html`

			`== text-horzline-with-bottom.html text-horzline.html`
			`== text-horzline-with-top.html text-horzline.html`

			`!= text-big-stroke.html text-blank.html`
			`!= text-big-stroke.html text-big-fill.html`

			`== text-context-state-test.html text-context-state-ref.html`
			`== text-font-inherit.html text-big-fill.html`
			`== text-space-replace-test.html text-space-replace-ref.html`

Fix crashes when using text API on a canvas context whose element doesn't have a frame, and update handling of that case to match the spec. (Bug 438695) r+sr=dbaron 2008-06-17 14:37:57 -07:00			`== text-no-frame-test.html text-no-frame-ref.html`
			`== text-no-frame-2-test.html text-not-in-doc-ref.html`
			`== text-not-in-doc-test.html text-not-in-doc-ref.html`

bug 581092 - use unrounded width when advancing x-position to draw RTL text run in canvas. r=vlad a=blocking2.0 2010-08-05 02:18:39 -07:00			`== text-bidi-ltr-test.html text-bidi-ltr-ref.html`
bug 698185 - test for failure to draw subsequent runs in mixed-dir canvas text. r=dbaron --HG-- rename : layout/reftests/canvas/text-bidi-ltr-test.html => layout/reftests/canvas/text-bidi-ltr-notref.html 2011-12-08 09:53:00 -08:00			`!= text-bidi-ltr-test.html text-bidi-ltr-notref.html # for bug 698185`
bug 581092 - use unrounded width when advancing x-position to draw RTL text run in canvas. r=vlad a=blocking2.0 2010-08-05 02:18:39 -07:00			`== text-bidi-rtl-test.html text-bidi-rtl-ref.html`
Canvas routines draw right-to-left text backwards - bug 402276 r=smontagu sr=roc 2008-07-18 11:29:06 -07:00
Bug 636753 - annotate reftest manifests to note tests that fail on Android (2 of 2). r=dbaron a=test-only 2011-03-18 08:54:15 -07:00			`fails-if(Android) != text-font-lang.html text-font-lang-notref.html`
Bug 478445 and bug 499628. strokeText needs to not re-stroke the current path. r=vlad, a=vlad 2010-11-30 10:19:56 -08:00
Test for bug 667947. r=dbaron 2011-06-28 17:35:39 -07:00			`== text-measure.html text-measure-ref.html`

Bug 478445 and bug 499628. strokeText needs to not re-stroke the current path. r=vlad, a=vlad 2010-11-30 10:19:56 -08:00			`== strokeText-path.html strokeText-path-ref.html`
Bug 524173. Reftests demonstrating off-by-one bug in degenerate linear gradients on some platforms. 2011-03-31 12:00:34 -07:00
bug 715798 pt 3.5 - canvas reftest to check that emoji character is non-blank. r=jrmuizel 2012-10-23 02:07:57 -07:00			`# check that emoji character renders as something non-blank (for Apple Color Emoji font, bug 715798)`
			`random-if(!cocoaWidget) fails-if(OSX==10.6) random-if(OSX==10.7) != text-emoji.html text-emoji-notref.html`

Bug 692879. Adjust reftests so that we pass with azure-quartz. r=mwoodrow 2011-11-25 16:38:14 -08:00			`# azure quartz uses CGDrawLinearGradient instead of DrawShading`
			`# so we have less control over degenerate behaviour as tested by this`
			`# test`
			`fails-if(azureQuartz) == linear-gradient-1a.html linear-gradient-1-ref.html`

No bug because we seem not to care, mark linear-gradient-1b.html as failing on 10.8+Azure too 2012-09-04 08:31:09 -07:00			`# this passes with cairo on 10.7 and 10.8 but not with azure for reasons unknown`
Switch reftest conditions for OSX to use the OSX variable from bug 789771 instead of a regex --HG-- extra : rebase_source : f4123ff10d1df1ec4f6f2ad75b764416056cf271 2012-09-08 23:40:29 -07:00			`fails-if(OSX==10.6\|\|(azureQuartz&&(OSX==10.7\|\|OSX==10.8))) == linear-gradient-1b.html linear-gradient-1-ref.html`
Bug 421865 - Handle zero dimensions correctly in canvas GetWidthHeight(). r=joe sr=roc 2011-05-19 03:49:18 -07:00
			`== zero-dimensions.html zero-dimensions-ref.html`
Bug 655926, Test even-odd fill rule. r=joe 2011-06-09 19:15:13 -07:00
			`== evenodd-fill-sanity.html data:text/html,<body>Pass`
			`!= evenodd-fill-1.html nonzero-fill-1.html`
			`== evenodd-fill-1.html evenodd-fill-ref.html`
			`== evenodd-fill-2.html evenodd-fill-ref.html`
			`== evenodd-fill-3.html nonzero-fill-2.html`
Bug 664884: Test mozCurrentTransform/mozCurrentTransformInverse. r=joe,roc 2011-06-29 14:34:58 -07:00
Bug 662038: Very basic tests of mozDash/mozDashOffset. r=joe 2011-06-29 14:34:58 -07:00			`== dash-sanity.html data:text/html,<body>Pass`
Bug 761890 (was 746883); change tests to work with Azure/Skia (mostly fuzz). r=gw280,jmuizelaar 2012-04-15 12:49:56 -07:00			`fuzzy-if(azureSkia,9,470) random-if(Android) == dash-1.html dash-1-ref.svg # Bug 668412 (really is android-specific, not IPC-specific)`
Bug 662038: Very basic tests of mozDash/mozDashOffset. r=joe 2011-06-29 14:34:58 -07:00
Bug 664884: Test mozCurrentTransform/mozCurrentTransformInverse. r=joe,roc 2011-06-29 14:34:58 -07:00			`== ctm-sanity.html data:text/html,<body>Pass`
			`fails == ctm-singular-sanity.html data:text/html,<body>Pass # Bug 612033`
			`== ctm-1.html ctm-1-ref.html`
Bug 672646. D2D: Fix trasparent radial gradients. r=roc The output merger stage was expecting premultiplied alpha and we were giving non-premultiplied. Fix by premultiplying in the shader. Also adds a reftest for this situation. 2011-07-21 15:40:41 -07:00
Switch reftest conditions for OSX to use the OSX variable from bug 789771 instead of a regex --HG-- extra : rebase_source : f4123ff10d1df1ec4f6f2ad75b764416056cf271 2012-09-08 23:40:29 -07:00			`fails-if(OSX==10.6) == 672646-alpha-radial-gradient.html 672646-alpha-radial-gradient-ref.html # Bug 673333`
Bug 674003. Add reftest for transparent radial gradients. r=joe This draws a unchanging part of the gradient on to another canvas. --HG-- extra : rebase_source : 57de827713cac43cf117da1d4dc95a29d0007987 2011-08-05 14:10:09 -07:00			`== 674003-alpha-radial-gradient-superlum.html 674003-alpha-radial-gradient-superlum-ref.html`
bug 693610 - reftest - ensure chinese and latin glyph runs are not overprinted. r=bas 2011-10-12 14:29:45 -07:00
			`!= 693610-1.html 693610-1-notref.html # bug 693610: multiple glyph runs should not be overprinted`

Bug 726951 - Reftest: Test for shadow clips working correctly in non-rectangular canvases. r=jrmuizel 2012-02-15 19:08:12 -08:00			`== 726951-shadow-clips.html 726951-shadow-clips-ref.html`

Bug 720721. Properly transform the clip path in DrawTargetCG. r=mwoodrow This is a little tricky because we can't put a save/restore pair around the operation because then we'll loose the clip. 2012-01-24 12:14:50 -08:00			`== transformed-clip.html transformed-clip-ref.html`
Bug 728028. CG: Avoid double transforming filled gradients. r=mwoodrow When we fixed transformed clips it caused us to double transform gradients. We fix this by avoiding ::PushClip when drawing gradients. This has the advantage of saving a save and restore pair and it makes Fill() more closely match Stroke() 2012-02-24 21:35:39 -08:00			`== transformed-gradient.html transformed-gradient-ref.html`
Bug 729850. CG: Get the point order correct when transforming a path. r=mwoodrow 2012-02-24 21:35:41 -08:00			`== transformed-path.html transformed-path.html`
Bug 749467 - Reftest: Add reftest for path continuation after transform changes. r=roc 2012-04-26 22:07:25 -07:00
			`== 749467-1.html 749467-1-ref.html`
Bug 784573 - Reftest for azure/cairo transform problem. r=joe 2012-09-03 14:49:33 -07:00
			`# You get a little bit of rounding fuzz on OSX from transforming the paths between user space and device space`
			`fuzzy-if(azureQuartz,2,113) fuzzy-if(d2d,12,17) == 784573-1.html 784573-1-ref.html`
Bug 802658 - Part 2: Apply matrix transform to text bounds; r=Bas 2012-10-28 13:52:21 -07:00
			`== 802658-1.html 802658-1-ref.html`